CA2155953A1 - Protein conjugates, compositions containing them and their applications as medicament - Google Patents

Abstract

Protein conjugates with specific affinity to appropriate target cells, and more particularly to cancer cells, as well as their applications as anti-cancer medicament(cytolytic properties). Said protein conjugates of the type comprising a non specific portion and a portion having a particular affinity for a specific type of target cells, are characterized in that the portion having such particular affinity for a specific type of target cells is .alpha.-foetoprotein, and the non specific portion of said conjugate is selected in the group including the cytotoxic substances selected from animal toxins, plant toxins, cytolytic enzymes and low molecular weight active anti-cancer principles.

Description

215595~

PROTEIN CONJUGATES, COMPOSITIONS CONTAINING THEM AND
THEIR APPLICATIONS AS DRUGS

The present invention relates to protein conjugates with specific affinity towards appropriate target cells, and more particularly towards cancer cells, as well as to their applications as anti-cancer drugs (cytolytic properties).
The transport of active molecules to specific biological sites was suggested by Erlich more than 100 years ago. The molecules used are more effective and have less pronounced side effects when their activity is manifested only on contact with the target organs or cells. This is particularly important in carcinology, where toxic or highly antigenic substances are administered.
The greatest progress in this field was obtained with molecules possessing both a non-specific portion and a portion possessing a particular affinity for a specific type of cell.
Since the development of monoclonal antibodies (mca), one of the aims of biotechnology has been the synthesis of immunotoxins, that is to say mca's specific for cells of one type, which are coupled to an agent capable of destroying the said cells. In general, an mca is prepared so as to have a specificity for an antigen present at the surface of the target cell, such that the mca will bind to the target cell, and, when it is incorporated into the said cell, it will transport its toxic conjugate into the cell.
The toxic conjugate is usually ricin or the subunit A of ricin. The subunit B binds to the galactose residues which are widely present at the cell surface, while the subunit A enzymatically inactivates ribosomes. It has been estimated that a single ricin molecule can inactivate all the ribosomes in a cell.
The invaded cell then dies, since it is incapable of producing new proteins.

215595~

Besides ricin, other agents may be incorporated into the immunotoxins; there may be mentioned, in particular, bacterial or plant toxins (or the active fragments thereof), radioactive atoms, and agents used in anti-cancer chemotherapy.
In all cases, the immunotoxins thus formed direct the active substance (the toxin or the anti-cancer agent) to the pathological area and thereby lead to the removal of the target cells.
The specific toxin/antibody conjugates form very active immunotoxins, but they also possess a non-specific toxicity which restricts their application in vivo.
Moreover, toxin fragments, which lack sites capable of recognizing the intended targets and conjugated to antibodies, are commonly employed- as immunotoxins. Such hybrid molecules such as, for example, those which contain a deglycosylated ricin-A
chain, are thus rendered highly specific. However, their activity is lower than that of immunotoxins comprising complete ricin and this activity depends on the intensity at which endocytosis takes place.
Indeed, translocation of the immunotoxin into the target cell depends essentially on the surface antigen recognized by the antibody component of the immunotoxin.
This is an important step in the action of immunotoxins which, in a great many cases, does not lead to endocytosis: consequently, intracellular translocation of the immunotoxin does not therefore take place. In such cases, the treatment thus results in failure. Furthermore, many antigenic labels, present at the cell surface, are also secreted into the extracellular medium. The effectiveness of the immunotoxin is consequently very much reduced on account of the effect of competition between the surface antigens and the secreted antigens.
In the particular case of antitumour therapy, the results obtained using immunotoxins consisting of 21~595~

cytotoxic substances linked to tumour-specific antibodies did not prove to be as conclusive as might be hoped from theory. Indeed, this theory assumed that these antibodies must have conducted the cytotoxic molecule into the cell. However, it is apparent that either the effectiveness of the cytotoxic molecule is lessened, or the affinity of the antibody is lessened, or both.
Moreover, the penetration of the cytotoxic substances into the target cells often proved to be very problematic and the cytotoxic effect thus very much reduced.
Furthermore, it has been discovered that many products obtained by conjugation between antibodies and such cytotoxic molecules cannot be used as drugs for reasons of solubility or diverse incompatibilities.
Among other difficulties, the choice of cytotoxic agent is often problematic. Indeed, this agent should be relatively non-toxic once conjugated and should become much more toxic after it has been transported to its site of action, where intracellular enzymes release the toxic substance into the tumour cells.
The objective consequently set forth by the Applicant was to have available a composition, of therapeutic aim, capable of acting only on specific targets (cancer cells) and capable of promoting endocytosis (stimulation of the transport of substances across cell membranes).
The subject of the present invention is protein conjugates, of the type comprising a non-specific portion and a portion possessing a particular affinity for target cells of a specific type, characterized in that the portion possessing a particular affinity for target cells of a specific type (also referred to as substance for specific transport to the target cell) is ~-foetoprotein, and in that the non-specific portion of the said conjugate is chosen from the group which comprises cytotoxic substances chosen from animal ~ - 4 -toxins, plant toxins, cytolytic enzymes and low molecular weight anti-cancer active principles and are selected from carboxyphosphamide, amboclorine, doxorubicin, bleomycin, cisplatin, vinblastine, calichemicine and methotrexate or substances which modify the metabolism of cancer cells, such as nucleotide sequences, which express a protein and the anti-sense nucleotide sequences.
Surprisingly, such conjugates retain some of the properties of ~-foetoprotein and are, in particular, readily incorporated into cells which carry the Q'- foetoprotein receptor on their surface, by a mechanism of endocytosis which stimulates penetration of the non-specific portion (cytotoxic substance, nucleotide sequence) and are only active towards cancerous cell lines such as lymphomas, hepatomas, neuroblastomas, melanomas, astrocytomas, terato-blastomas and carcinomas of embryonic, ovarian, testicular or presacral type.
Equally surprisingly, besides the fact that the non-specific portion (cytotoxic substance, enzyme, anti-cancer active principle) and Q~-foetoprotein retain their effectiveness, the conjugate has an activity with respect to target cells which is significantly higher than that obtained with the non-conjugated non-specific portion.
Another subject of the present invention is pharmaceutical compositions, characterized in that they comprise, as active principle, at least one protein conjugate in accordance with the invention and at least one pharmaceutically acceptable vehicle.
The ~-foetoprotein - cytotoxic substance protein conjugates chosen from anti-cancer active principles are particularly able to be used for the preparation of a drug intended for use in the treatment of cancers and especially in the treatment of leukaemias, lymphomas, hepatomas, neuroblastomas, melanomas and astrocytomas.

_ ~ - 5 -The ~-foetoprotein-vinblastine, ~-foetoprotein-doxorubicin and ~-foetoprotein-calichemicine conjugates are preferably particularly able to be used for the preparation of a drug intended for use in chronic myeloid leukaemia.
Apart from the preceding arrangements, the invention also comprises other arrangements, which will emerge from the description which follows, referring to examples of implementation of the process which forms the subject of the present invention.
It should, however, be understood that these examples are given solely by way of illustration of the subject of the invention, of which they in no way constitute a limitation.
EXAMPLE 1: Production of ~-foetoprotein a) Preparation:
Amounts of ~-foetoprotein of the order of a milligram may be prepared by immunoaffinity techniques.
(1) The first step consists therefore in preparing this protein in a pure enough state and in sufficient quantity (about 100 ~g) to immunize mice and prepare monoclonal antibodies which will then be used to purify the ~-foetoprotein in large amounts in a single step.
~-Foetoprotein may be prepared from blood of the human umbilical cord or from cancer cells in culture (hepatomas, teratocarcinoma). The method of production consists, in effect, in separating this protein from the other proteins which accompany it, according to methods which are standard and known per se, such as chromatography by ion exchange, by affinity, ~y exclusion on gel or by hydrophobic interaction, preparative electrophoresis, dialysis and ultrafiltration.
In a first step, the starting material, whole blood or cell suspension, is centrifuged and then subjected to a dialysis or an ultrafiltration in order to remove mineral salts and low molecular weight molecules. The total proteins are then purified by chromatography followed by preparative electrophoresis, so as to avoid any contamination by foreign proteins and in particular by albumin.
More precisely, 10 ml of human foetal umbilical serum (12 weeks of gestation) are dialysed against a pH 7.0 buffer solution (6.5 mM bis-tris propane), for 24 hours, through a regenerated cellulose membrane with a cutoff threshold of 6000-8000 daltons. The dialysed serum is then placed on a Cibacron Blue 2-Sepharose~
CL-6B column (Pharmacia) 26 mm in diameter and 70 cm in length, in the same buffer. The aim of this operation is to remove the majority of the albumin which remains bound to the gel.
The eluted proteins are collected and chromatography is then performed on a Mono Q column (Pharmacia FPLC System). The elution is carried out by a gradient of sodium chloride (0-0.5 M) in the same buffer as above. The ~-foetoprotein is eluted by 0.35 M
sodium chloride. After concentration by ultrafiltration on PM-10 membrane with a cutoff threshold of 10,000 daltons (Amicon), the proteins are again chromatographed by filtration on crosslinked agarose gel (Superose~ 12, Pharmacia), in a column 25 mm in diameter and 80 cm in length in a pH 7.0 phosphate buffer. After having standardized the column, the peaks corresponding to molecular weights of 70 kDa and 140 kDa are collected. These two fractions are combined and then subjected to preparative electrophoresis on polyacrylamide gel: the electrophoresis is carried out without SDS, starting from a zone of concentration gel (4% polyacrylamide in the concentration gel and 7.5% in the separation gel). A Tris-HCl buffer pH 6.7 is used in the concentration gel and a Tris-glycine buffer pH 8.3 is used in the separation gel.
The sample is diluted by half with Tris-HCl buffer pH 6.7. One part of glycerol is added per one part of sample diluted in the buffer, and one-tenth of the volume to be deposited of bromophenol blue solution in TrisHCl buffer pH 6.7 is added, in order to 21559~3 visualize the migration front. The electrophoresis is carried out at 15C under 30 mA for about four hours.
In order to determine the location of the ~-foetoprotein on the gel after migration, an electrophoresis of foetal serum is first carried out against adult serum under the same conditions. The concentration of ~-foetoprotein in adult blood is extremely low. Thus, comparison of the two electrophoreses makes it possible to locate the band corresponding to ~-foetoprotein between the albumin and transferrin in foetal blood.
The gel is stained in part with Coomassie blue and the band, in its non-stained part corresponding to the a-foetoprotein, is carefully cut out in order to avoid contamination with the albumin still present at this stage in the purification. The gel containing the ~-foetoprotein is homogenized in pH 7.0 phosphate buffer and subjected to agitation in a tube for 1 hour.
After centrifugation, the supernatant is collected and chromatographed on Cibacron Blue 2-Sepharose~ as above, but in pH 7.0 phosphate buffer.
The proteins not retained by the gel are collected and then concentrated by ultrafiltration, as above. These proteins are then used to immunize mice in order to obtain antibodies which will allow significant amounts of ~-foetoprotein to be prepared by immunoaffinity.
(2) At this stage, the ~-foetoprotein is used to produce antibodies of low anti-~-foetoprotein affinity. This production takes place by injecting ~-foetoprotein into mice, in the presence of Freund adjuvant. The popliteal lymphatic ganglions are then removed from these mice and the cells of these ganglions are fused with myeloma cells in order to form hybridomas. The period of immunization is deliberately chosen to be very brief (one week), in contrast with the usual methods which last for several months.
Moreover, ganglion cells are used here in order to carry out the fusion instead of splenocytes.

~5~953 This method makes it possible to produce clones of anti-~-foetoprotein antibodies of low affinity.
More precisely, two approximately four-month-old BALBtC mice are immunized by injection of 20 ~g of proteins combined with Freund adjuvant, into a hind foot, twice with a two-day interval, and then a third time with 50 ~g of adjuvant-free proteins two days later.
Three days after the last injection, i.e. nine days in total, the mice are sacrificed and the popliteal lymphatic ganglions are removed. They are homogenized and washed with DMEM medium containing no foetal calf serum, and the homogenate is used to carry out a fusion with Sp2/o myeloma cells. These mouse myeloma cells are conventionally used to produce hybridomas; they have the property of not growing in HAT medium.
The Sp2/o cells are cultured on DMEM-Hybrimax medium with 10~ foetal calf serum. The fusion is carried out in the presence of PEG 3000, at a concentration of 50~ in DMEM medium from Fl thymocytes (CBA x BALB/c). The hybridomas are cultured in HAT
medium. The clones appear in 10-15 days in 75~ of the cultures and correspond to Sp2/o complemented with immunocompetent cells.
About 40 different clones produce anti-~-foetoprotein monoclonal antibodies. 9 of them do not react with normal human serum by the ELISA technique and 5 do not react by the method of immunological analysis after transfer (immunoblotting). Two clones are chosen in order to separate the ~-foetoprotein in larger amounts; they possess no cross-affinity with other serum proteins.
to 50 million hybridoma cells are then injected into a mouse in order to cause it to develop a tumour producing monoclonal antibodies. Between the 10th and 17th day after the implantation, the ascitic fluid is removed in order to recover the antibodies.

g The monoclonal antibodies of IGg type are then purified by chromatography on Protein-A-Sepharose~
(Pharmacia).
5 ml of ascitic fluid are centrifuged at 3000 rpm for 20 minutes, diluted by half with Tris-HCl buffer pH 8.1 and placed on a Protein-A-Sepharose~
column. The column is washed with 100 ml of Tris-HCl buffer pH 8.1 and the immunoglobulins are eluted with pH 6.0 acetate buffer, then pH 4.5 and finally pH 3.5.
The antibodies with low anti-~-foetoprotein affinity of one of the clones (lOC3) are eluted at pH 4.5, and another (7H8) at pH 3.5.
The immunoglobulins are then concentrated by precipitation with 50~ ammonium sulphate, at 4C, for 18 hours and are centrifuged for 30 minutes at 4000 rpm. The centrifugation pellet is redissolved in a minimum volume of PBS buffer and dialysed against the same buffer, for 24 hours. 3 to 5 mg of antibody are thus obtained per 1 ml of ascitic fluid removed from the mice.
The immunoglobulins thus purified are then immobilized on a Sepharose~ gel (Pharmacia) activated with cyanogen bromide according to the method recommended by the gel manufacturer.

(3) Preparation of the ~-foetoprotein in larger amounts may then be undertaken by immunoaffinity chromatography after the antibodies have been immobilized on a support such as Sepharose~ activated with protein A, for example, so as to read the fraction Fc of Igl, Ig2 and Ig4.
This general method makes it possible to obtain substantial amounts of ~-foetoprotein in high purity.
When prepared under these conditions, this protein proves to be stable.
More precisely, a column 25 mm in diameter and 50 cm in length is prepared with Sepharose gel on which the monoclonal antibody as described above is bound.
The column is loaded with 50 ml of human foetal serum and the ~-foetoprotein is eluted with a phosphate 21~5953 buffer in the form of a gradient of pH ranging from 3.5 to 6Ø The ~-foetoprotein is eluted at pH 6.0, in a proportion of about 90~.
b) Conjugation of ~-foetoprotein with various active molecules:
~ -Foetoprotein is a glycoprotein. Accordingly, the two domains of the molecule, proteinic and osidic, may be used as binding sites for various substances.
Any substance having a chemically active site may, in principle, be conjugated with ~-foetoprotein. In a conventional manner, molecules containing thiol residues may be linked to ~-foetoprotein by a reaction between amino groups of the protein and activated carboxylic esters or with imidates. Molecules containing carboxylic residues are bound particularly effectively by carbodiimide derivatives. In general, the binding to osidic residues is carried out by modification with sodium metaperiodate, especially when this involves binding molecules containing free amino residues.
All these processes are already known per se.
It may thus be envisaged to bind a large number of substances, such as cytotoxic molecules, genetic material (nucleotide sequences), inhibitors of genomic replication (anti-sense DNA for example), enzymes or alternatively fluorescent agents, with the aim of locating cancer cells under ultraviolet light.
c) Targets of conjugated ~-foetoprotein:
Any cell having the ~-foetoprotein receptor on its surface is a possible target for conjugated ~-foetoprotein. This receptor is not present on any normal or non-foetal cell. Only foetal cells and cancer cells have a proportion thereof which exceeds 90~, in more than 70 different cases of cancers (R. Moro et al., Tumor Biol., 8, 293, 1987).

~1~59~3 EXAMPLE 2: Preparation of conjugates in accordance with the invention a) a-Foetoprotein-toxin conjugates - a-foetoprotein (AFP) and diphtheria toxin (DT).
To 10 ml of an ~-foetoprotein solution concentrated to 1 mg/ml in PBS buffer are added 100 mg of SPDP (N-succinimidyl 3-(2-pyridyldithio)propionate) dissolved in 160 ml of dimethylformamide. The solution is then dialysed against PBS buffer or chromatographed on Sephadex~ G-25 (Pharmacia), after equilibrating with PBS buffer. The protein thus modified is collected and concentrated by ultrafiltration on a cellulose membrane having a cutoff threshold of 10,000 daltons.
10 mg of diphtheria toxin are placed in an aqueous solution containing 50 mM of dithiothreitol and are incubated for 1 hour at room temperature. The protein thus reduced is then purified by chromatography on a Sephadex~ G-25 column equilibrated in PBS buffer, or dialysed against the same buffer. The reduced toxin is placed in the presence of the modified a-foetoprotein and the solution of the two proteins is concentrated by ultrafiltration on a cellulose membrane having a cutoff threshold of 10,000 daltons. The mixture is then left to incubate for 18 hours at room temperature.
The product obtained is then purified by chromatography on a Sephacryl~ S-3.0 column (Pharmacia).
The final yield for the conjugation is 34~.
- a-foetoprotein (AFP) and ricin A (R).
1 mg of a-foetoprotein is dissolved in 1 ml of PBS buffer and mixed with 6 mg of SPDP (N-succinimidyl 3-(2-pyridyldithio)propionate) in ethanolic solution to a proportion of 1 mg/ml. The mixture is incubated for minutes, at room temperature, with constant stirring. The a-foetoprotein thus modified is dialysed against a 0.05 M borate buffer, pH 8.5.

~155953 In parallel, ricin A is treated under the same conditions and in the same amounts as the ~-foetoprotein.
The two modified proteins are then placed together, concentrated by ultrafiltration on a cellulose membrane with a cutoff threshold of 10,000 daltons down to a final volume of 5 ml, and are incubated for 18 hours at room temperature.
The resulting product from the conjugation thus performed is then purified by filtration chromatography on Sephacryl~ S-300 gel. The conjugate is recovered in its pH 8.5 buffer, at the molecular weight corresponding to the sum of the molecular weights of the two proteins, i.e. about 102,000 daltons.
The final yield for the conjugation is 42~.
b) ~-Foetoprotein-enzyme conjugates - ~-foetoprotein (AFP) and asparaginase (Asp).
In order to perform the said conjugation, 1-cyclohexyl-3-(2-morpholinoethyl)carbodiimide (CMC) is used. Human AFP (Asp:AFP molar ratio of 1:1) is added to a solution of asparaginase, activated by an excess, of a factor of the order of 2000, of CMC, at pH 5.4.
The mixture obtained is incubated for 16 hours at room temperature and dialysed against PBS.
c) ~-Foetoprotein - low molecular weight biologically active substance conjugates - ~-foetoprotein (AFP) and carboxyphosphamide (CFA).
Carboxyphosphamide is the metabolite of cyclophosphamide, a widely used anti-cancer molecule.
Its metabolite, carboxyphosphamide, cannot penetrate into cells on account of its high negative charge, and thus exhibits a low toxicity. If, by means of conjugation, carboxyphosphamide penetrates into the cells, it will be cleaved into acrolein and phosphoramide by phosphoamidases which are particularly active in cells during proliferation. The phosphoramide thus obtained is highly cytotoxic.

..

The conjugation between ~-foetoprotein and carboxyphosphamide is carried out at +4C in pyridine containing 0.01 M HCl and EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide) in an EDC:AFP
proportion of 2500:1. Furthermore, hexamethylene-diamine (HMDA) is used as intercalating agent. The final a-foetoprotein-hexamethylene diamine-CFA pro-portions are 1:10:120. The conjugate which results therefrom is dialysed against PBS and then stored at -70C.
The final yield for the conjugation is 72%.
- ~-foetoprotein (AFP) and amboclorine (AC) .
Amboclorine, of formula HOOC-(CH2)3-C6H4-N(CH2-CH2-Cl)2, may be conjugated with ~-foetoprotein in the same manner as carboxyphosphamide, in a yield of about 65%.
- ~-foetoprotein (AFP) and methotrexate (MT).
Methotrexate (MT) is dissolved in a pyridine-HCl buffer, pH 5.0 and the ~-foetoprotein (AFP) solution is prepared in the same buffer. EDC is added and the whole mixture is stirred. The MT:AFP:EDC ratio is 10:1:2500. After stirring for 3 hours, the preparation is dialysed against a PBS buffer and diluted in PBS until an AFP concentration of 20 mg/ml is obtained.
a-foetoprotein (AFP) and doxorubicin (DR).
DR is dissolved in a pyridine-HCl buffer, pH 5.0 and the ~-foetoprotein (AFP) solution is prepared in the same buffer. EDC and adipic acid are added. Adipic acid acts as an intercalating agent between the AFP and the DR; the DR:adipic acid:AFP:EDC
ratio is 10:50:1:2500.
- Conjugation of ~-foetoprotein with daunomycin.
Daunomycin has a free amino radical, which allows direct conjugation with ~-foetoprotein.
Daunomycin is dissolved in pyridine to a concentration of 10%. A solution of ~-foetoprotein in a pH 5.0 hydrochloric buffer is added, followed by 2155~53 addition of adipic acid as intercalating agent so that the proportions between the various products are variable but the ratio between ~-foetoprotein and EDC
is still 1 to 2500.
The yield for the conjugation very much depends on the various proportions between the products and ranges from 22~ to 57~.
- ~-foetoprotein (AFP) and rubomycin (RM).
The AFP-RM conjugates are prepared according to the method used for doxorubicin.
- a-foetoprotein (AFP) and bleomycin (BM).
The AFP-BM conjugates are prepared according to a method similar to that used for doxorubicin (no addition of adipic acid); the DM:AFP:EDC ratio is 100:1:2500.
- ~-foetoprotein (AFP) and cisplatin (CP).
AFP is added to a solution of cisplatin and EDC
in a 0.1 M pyridine - HCl buffer, pH 5Ø The reaction mixture is stirred overnight at 4C or for 2 hours at 20C. The conjugate obtained is dialysed 3 times against a PBS buffer at 4C, for 24 hours. Samples of conjugate may be stored in a freezer. The EDC:AFP molar ratio should not be less than 2000:1.
- ~-foetoprotein (AFP) and vinblastine (VB).
Vinblastine has no free amino group;
consequently, it is necessary to esterify it before performing a conjugation with AFP.
The esterification is performed, at 37C, by dropwise addition of 0.1 M KOH to a solution of vinblastine, until a pH of 9.5 is obtained. The reaction mixture is neutralized with 0.1 M pyridine -HCl solution at pH 5Ø The vinblastine thus modified is conjugated with AFP, according to a method similar to that described above for cisplatin; however, for the application of this method to vinblastine, the EDC is introduced last, since vinblastine interacts with the amino groups of AFP.

REPLACEMENT SHEET (RULE 26) 2~5953 - a-foetoprotein (AFP) and calichemicine (CCh).
In order to obtain better conjugation conditions, calichemicine is first cleaved, by addition of dithiothreitol (up to 1 M) and the AFP is modified with SPDP in PBS, pH 7.2, at 4C, for 12 hours. The solutions of calichemicine, of EDC and of modified AFP
are mixed together with stirring overnight. The conjugate obtained is dialysed against a PBS buffer, for 24 hours. The SPDP:AFP molar ratio depends on the desired calichemicine:AFP ratio, and should be greater than the latter ratio by a factor of 8.
EXAMPLE 3: Effects of the various conjugates in accordance with the invention on h~an cancer cell lines in culture.
The action of the conjugates was tested on the following lines;
lymphoma T QOS (QOS), lymphoma T CEM (CEM), lymphoma B Raji (Raji), lymphoma B Namalva (Nam), hepatoma HepG2a (Hep), astrocytoma (Ast), melanoma Bro (Bro), and neuroblastoma IMR - 32 (IMR).
The experimental procedure, common to all the tests carried out, is as follows: the cells are cultured in plates containing 96 wells (Linbro~) in an RPMI medium complemented with 10~ foetal calf serum, with penicillin (100 units/ml) and with streptomycin (100 mg/ml), at 37C and under an atmosphere containing 5~ of CO2. The conjugate tested is added to a concentration of 5 mg/ml and, after 48 to 72 hours, the proliferating activity is checked.
The abovementioned culture medium may, in addition, be adapted to each specific culture.
The results obtained are illustrated in Tables I to XI below, which give the percentage of surviving cells after 72 hours of incubation with the cytotoxic substance alone (control) or with a conjugate in accordance with the invention (test).
- ~-foetoprotein (AFP) and diphtheria toxin (DT).
The results are illustrated in Table I below:

REPLACEMENT SHEET (RULE 26) TA~3LE I

Effect of the AFP-TD conjugate on human T QOS lymphoma cell lines (~ of surviving cells) Test Control 20 ~g/ml AFP and 48 58 12 ~g/ml DT

20 ~g/ml AFP and 44 34 24 ~g/ml DT
Control: test performed with DT alone, at the same concentrations.
- a-foetoprotein (AFP) and ricin A (R).
The results are illustrated in Table II below:
TABLE II
Effect of the AFP-R conjugate on human T QOS lymphomas (~ of surviving cells) Test Control 20 ~g/ml AFP and 51 62 10 ~g/ml R

20 ~g/ml AFP and 14 20 20 ~g/ml R
Control: test performed with R alone, at the same concentrations.
- ~-foetoprotein (AFP) and carboxyphosphamide (CFA).
A) The cytotoxicity of ~-foetoprotein coupled to carboxyphosphamide is determined with variable HMDA-~-foetoprotein proportions of between 0 and 120.
This determination is performed on QOS cells, belonging to a lymphoblastoid line, by measuring the amount of tritiated thymidine incorporated relative to the controls (~-foetoprotein (AFP) alone, ~_ ~15595~

carboxyphosphamide (CPA) and HMDA alone or as a mixture in the same proportions as in the conjugate).
1. Without HMDA (% of tritiated thymidine incorporated):

Concentration CPA:AFP molar Effect Effect of CPA in the proportion CPA-AFP CPA
medium 0.75 ~M 10:1 25% 0%
1.50 ~M 20:1 98% 0~
3.75 ~M 50:1 100% 0%
9.00 ~M 120:1 100% 2%
2. With HMDA (% of tritiated thymidine incorporated):

Concentration CPA:HMDA:AFP Effect Effect of CPA in the molar CPA-HMDA CPA+AFP
medium proportion AFP

0.75 ~M 120:20:1 97~ 0%
1.50 ~M 50:20:1 100% 0%
3.75 ~M 20:50:1 100% 0%
9.00 ~M 10:120:1 100% 0%
b) The following results are obtained under the same conditions as those illustrated in Tables I and II
(% of surviving cells):
TABLE III

AFP-CFA CFA AFP
1:120 (0. 8 ~M) Effect 1 87 144 - ~-foetoprotein (AFP) and amboclorine (AC).
The results are illustrated in Table IV below:

REPLACEMENT SHEET (RULE 26) ~1559~3 , TABLE IV
Effect of the AFP-AC conjugate on various human cell lines (~ of surviving cells) Type AC AFP:AC AC AFP:AC AC AFP:AC
of (60 nM) = 1:5 (120 nM) = 1:10 (240 nM) = 1:20 cell (60 nM) (120 nM) (240 nM) Hep 98 97 89 81 80 61 Bro 89 87 89 65 73 48 - ~-foetoprotein (AFP) and methotrexate (MT).
a) Comparative tests are conducted on QOS cells (lymphoblastoid line) and normal human lymphocytes. The conjugates, comprising various proportions of methotrexate, are added to cultures of QOS cells and to cultures of normal human lymphocytes in boxes containing 96 wells (200 ~l of medium). The cultures are incubated for 72 hours. The living cells are detected and counted by staining with thiazolyl blue (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide).
A solution of thiazolyl blue at a concentration of 1 mg/ml is added at the end of the incubation, to a proportion of 50 ~l per well. Incubation is again carried out for 16 hours, followed by centrifugation for 5 minutes at 5000 rev/min. The supernatant is removed (150 ~l accurately measured) and 150 ~l of DMSO
are added. After complete dissolution of the crystals of formazan, the O.D. is read at 540 nm. The conjugate-free cultured control is assigned a value of 100~.

~155~.~3 Concentration Methotrexate- % of % of of methotrexate AFP surviving surviving in the mediumProportion QOS cells lymphocytes 51.50 ~M 50:1 60 87 3.00 ~M 100:1 53 82 6.00 ~M 200:1 48 91 12.00 ~M 400:1 45 90 b) The effect of the AFP-MT conjugate is also studied on other cell lines. The results are illustrated in Table V below:
TABLE V
Effect of the AFP-MT conjugate on human cell lines (~ of surviving cells) TypeMT AFP:MT MT AFP:MT MT AFP:MT
of(24 nM) = 1:2(60 nM) = 1:5 (240 nM) = 1:20 cell (24 nM) (60 nM) (240 nM) Raji91 7680 72 Nam 88 29 75 56 57 52 Bro 98 81 90 55 61 50 Ast 93 90 83 41 40 35 - ~-foetoprotein (AFP) and doxorubicin (DR).
a) Comparative tests are conducted on QOS cells (lymphoblastoid line) and normal human lymphocytes exactly as for methotrexate.

REPLACEMENT SHEET (RULE 26) 2155~9~3 , Concentration Doxorubicin- ~ of ~ of of methotrexate AFP surviving surviving in the mediumProportion QOS cells lymphocytes 51.50 ~M 50:1 22 66 3.00 ~M 100:1 6 74 6.00 ~M 200:1 3 56 12.00 ~M 400:1 2 71 b) The effects of the AFP-DR conjugate have also been studied on other cell lines; the results are illustrated in Table VI below:
TABLE VI
Effect of the AFP-DR conjugate on human cell lines (~ of surviving cells) TypeDR AFP:DR DR AFP:DR DR AFP:DR
of(24 nM) = 1:2(48 nM) = 1:4 (60 nM) = 1:5 cell (24 nM) (48 nM) (60 nM) Raji99 23 Nam 90 2581 6 65 3 Bro 91 5595 80 71 36 Hep 104 8195 80 71 30 Ast - - 87 70 55 14 - ~-foetoprotein (AFP) and daunomycin (DN).
The test is performed on the same cells and under the same conditions as for ~-foetoprotein conjugated with carboxyphosphamide (Test a).

REPLACEMENT SHEET (RULE 26) ~ ~ 5 5 9 ~ 3 , 1. Without intercalatinq aqent:

Concentration DN:AFP Effect Effect of DN in the molar DN-AFP DN
medium proportion 0.15 ~M 10:1 75~ 5 0.30 ~M 20:1 75~ 26 0.75 ~M 50:1 100~ 37~
1.50 ~M 100:1 100~ 100%
1.80 ~M 120:1 100~ 100 2. With adipic acid as intercalatinq aqent:

Concentration DN:AFP Effect Effect of DN in the molar DN-AFP DN
medium proportion 0.15 ~M 10:50:1 75~ 5 0.75 ~M 50:120:1100~ 37 1.80 ~M 120:20:1100~ 100 1.80 ~M 120:50:1100~ 100 1.80 ~M 120:120:1100~ 100~
1.80 ~M 120:250:1100~ 100%
- ~-foetoprotein (AFP) and rubomycin (RM).
The effects obtained with the AFP-RM conjugate on human T QOS lymphoma cells are illustrated in Table VII below:
TA~3LE VII
Effect of the AFP-RM conjugate on human T QOS lymphomas ~ of surviving cells) 3 0 Type RM AFP: RM RM AFP: RM RM AFP: RM
of (24 nM) = 1:2 (60 nM) = 1:5 (120 nM) = 1:10 cell (24 nM) (60 nM) (120 nM) ~155953 - ~-foetoprotein (AFP) and bleomycin (BM).
The results are illustrated in Table VIII
below:
TABLE VIII
Effect of the AFP-BM conjugate on human cell lines (~ of surviving cells) Type of cell AFP:BM = 1:100 AFP BM
(120 nM) Hep 45 103 56 Ast 29 124 44 - ~-foetoprotein (AFP) and cisplatin (CP).
The results are illustrated in Table IX below:
TABLE IX
Effect of the AFP-CP conjugate on human cell lines (~ of surviving cells) TypeCP AFP: CP CP AFP: CP CP AFP: CP
of(24 nM) = 1:2(48 nM) = 1:4 (60 nM) = 1:5 cell (24 nM) (48 nM) (60 nM) Ast120 76108 60 98 48 Bro97 78 90 72 65 51 - ~-foetoprotein (AFP) and vinblastine (VB).
The results obtained on astrocytoma cells are illustrated in Table X below:

REPLACEMENT SHEET (RULE 26) ~155953 TABLE X
Effect of the AFP-VB conjugate on human astrocytoma cell lines (~ of surviving cells) Type VB AFP: VB VB AFP: VB VB AFP: VB
of(24 nM) = 1:2 (48 nM) = 1:4 (60 nM) = 1:5 cell (24 nM) (48 nM) (60 nM) Ast42 24 41 26 38 20 - ~-foetoprotein (AFP) and calichemicine (CCh).
The results obtained on various human cell lines are illustrated in Table XI below:
TABLE XI
Effect of the AFP-CCh conjugate on human cell lines (~ of surviving cells) Type CCh AFP CCh: AFP: CCh AFP: CCh AFP: CCh AFP:
of ~24 CCh (48 CCh (60 CCh (240 CCh(24 CCh cell pM) 1:2 pM) 1:4 pM) 1:5 pM) 1:2nM) 1:2 (24 (48 (60 (240 (24 2 0 pM) pM) Pm) pM) nM) Raji - - - - - - 58 4121 10 Bro 100 55 95 38 90 30 Incubation for 48 hours.
EXAMPLE 4: Inhibitory effects of the various conjugate~
in accordance with the invention on the proliferation of various leukaemia cells.
I - Effects of the AFP-DR and AFP-VB coniuqates on various leukaemias in man.
a) Equipment and method:
After withdrawing the blood (5-7 ml), the samples obtained are incubated at 37C for 30-45 min, in order to allow erythrocytes to sediment out. The plasma is removed and the amount of lymphocytes is evaluated. The plasma is transferred to an RPMI medium -~ 2~55~53 in the presence of various antibiotics (penicillin, gentamycin). The amount of lymphocytes should not exceed 2 x 106/ml. When there are large amounts of leukaemia cells in the sample, bovine foetal serum is added.
The proliferative activity of the cells is measured by a standard method on plates using tritiated thymidine after 24, 48 and 72 hours.
b) Patients:
Patient 1: chronic myeloid leukaemia, Philadelphia chromosome + (9-year-old child), Patient 2: chronic myeloid leukaemia, Philadelphia chromosome - (69-year-old woman), Patient 3: acute leukaemia (10-year-old girl), Patient 4: primary lymphosarcoma.
c) Results:
The results obtained with the AFP-DR and AFP-VB
conjugates, in the four abovementioned patients, are illustrated in Tables XII-XIV below:
TABLE XII: Patient 1 The tests, with Patient 1, were performed on lymphocytes and on plasma, after 24 and 48 hours.

Total Lymphocytes Total Lymphocytes plasma (24 h) plasma (48 h) (24 h) (48 h) DR 5.3~ 6.1% 7.0~ 8.1%

AFP:DR 1:50 2.9~ 2.2~ 1.8~ 3.9 VB 65.5~ 52.1 AFP:VB 1:10 48.3~ 43.8~
These results show that after incubation for 24 hours with the 1:50 AFP-DR conjugate, a decrease by a factor of the order of 35 is observed in the proliferative activity, whereas DR alone only results in a decrease by a factor of the order of 20 (total plasma).

REPLACEMENT SHEET (RULE 26) ~155~i3 . , , I

In the case of the lymphocytes, the inhibitory effects are respectively of a factor of the order of 45 and 16.
After incubation for 48 hours, the effect is even more pronounced.
The inhibitory effect of the AFP-VB conjugate proved to be less pronounced, but the test was performed with very low concentrations of VB.
The tests performed with Patient 2 were carried out solely on plasma.

TABLE XIII: Patient 2 Total plasma 24 h 48 h 72 h DR 58~ 70~ 33 AFP:DR 1:5047~ 55~ 21 VB 68~ 55~ 31%

AFP:VB 1:1053~ 37~ 17~
The inhibitory effect of the conjugates is less pronounced than for Patient 1; this is due to the type of leukaemia and to the patient's age and previous treatments.
The tests with Patients 3 and 4 were performed on whole blood, after incubating for 72 hours.

REPLACEMENT SHEET (RULE 26) 2L5595~

TABLE XIV

Patient 3 Patient 4 DR 14~ 78 AFP:DR = 1:50 7~ 48 CCh 29~ 92 AFP:CCh = 1:10 8~ 45 VB 10% 88 AFP:VB = 1:10 7~ 62~
For Patient 3, the inhibitory effect of the conjugates is significantly more pronounced than that obtained with DR or CCh, alone.
II - Effects of the AFP-CCh coniuqates on mouse leukaemias.
A P388 or L1210 leukaemia is transmitted by injection of 107 suitable cells to male DBA mice weighing about 18 g. On the second day after the injection, when the tumour has a size of about 0.01 cm3, the treatment is established and comprises the SC injection of an AFP-CCh conjugate (AFP:CCh ratio of 1:4), in accordance with the invention or of calichemicine alone (CCh at a dose of 0.6 ~g/kg).
The anti-cancer preparations are introduced, subcutaneously into the region of the tumour, or intravenously (once/day for 7 days). A first group of animals serving as a control received no preparation; a second group of animals also serving as a control received 0.2 ml of 0.9~ NaCl solution subcutaneously.
Each group is made up of 6 animals. The increase in lifespan is calculated according to the following formula: ~ILS=100. (T/C-1), in which T represents the average survival time in days of the experimental group and C represents the average survival time in days of the control group. The size of the tumour is also monitored. The volume of the tumour is calculated 21559~;~
, during the experiment, according to the following formula: V=1/2 L.W2, in which L represents the length of the tumour and W represents the width of the tumour.
The results are illustrated in Tables XV, XVI and XVII
below:
TABLE XV

Type Agent Route of Of Volume cm3 tumour admin.

0 3rd 5th 7th 9th day day day day day Control - 0.01 0.18 0.61 1.05 3.1 0.9~ s.c. 0.01 0.10 0.30 0.83 2.81 NaCl CCh i.v. 0.01 0.12 0.74 1.14 3.42 P388 CCh s.c. 0.01 0.08 0.19 0.42 1.21 AFP- i.v. 0.01 0.09 0.34 1.13 2.6 CCh AFP- s.c. 0.01 0.08 0.14 0.27 0.81 CCh control - 0.01 0.05 0.19 0.22 0.6 0.9% s.c. 0.01 0.09 0.15 0.31 0.56 NaCl CCh i.v. 0.01 0.02 O.lS 0.30 0.52 L1210 CCh s.c. 0.01 0.03 0.08 0.12 0.29 AFP- i.v. 0.01 0.04 0.11 0.20 0.28 CCh AFP- s.c. 0.01 0.03 0.10 0.09 0.16 CCh .

TABLE XVI

Leukaemia Agent Route of Life ~ ILS
admin. span Control - 20 0 O.9t NaCl s.c. 21 7 CCh i.v. 21.5 7.5 CCh s.c. 24.5 22.5 P388 AFP-CCh i.v. 19 -5 AFP-CCh s.c. 27.8 39 control - 11.5 0 PBS s.c. 13 13 CCh i.v. 12.5 8.7 L1210 CCh s.c. 17 48 AFP-CCh i.v. 14 21.7 AFP-CCh s.c. 19.5 69.5 REPLACEMENT SHEET (RULE 26) _ ` 2155953 TABLE XVII

Agent Dose number of Life ~ ILS
~g/kg mice span Control0.9~ NaCl 30 7 0 CCh 0.9 6 9.2 31.4 AFP-CCh 0.9 6 10.5 51.8 CCh 2.7 6 9.7 38.6 AFP-CCh 2.7 6 8.5 21 CCh 8 6 9.5 35.7 AFP-CCh 8 6 11.5 60.7 CCh 24 6 6 -14 AFP-CCh 24 6 5.7 -18.6 Table XV shows that, both in leukaemia P388 and in leukaemia L1210, a significant decrease in the volume of the tumours is observed in the animals treated. These results confirm the value of the conjugates in accordance with the invention. Table XVI
shows the value of the SC administration of a conjugate in accordance with the invention.
AS regards leukaemia L1210, the effect of various concentrations of calichemicine alone or in the form of a conjugate was tested (concentrations from O.9 llg to 24 ~g); Table XVI shows that the best results are obtained using low concentrations of cytotoxic agent.
III - Effects of the AFP-VB coniuqates on mouse leukaemia.
A P388 or L1210 leukaemia is transmitted by injection of 107 suitable cells to male DBA mice weighing about 18 g. On the second day after the injection, when the tumour is about 0.01 cm3 in size, the treatment is established and comprises the SC

21559^~3 ~, . . .

injection of V~3 or of AFP-V~3 conjugate (VL at a dose of 200 ~g/kg, the AFP:V~3 ratio being 1:10) in the region of the tumour (once/day for 7 days). A first group of animals serving as a control received no preparation; a second group of animals also serving as a control received 0.2 ml of 0.9~ NaCl solution subcutaneously.
Each group is made up of 6 animals. The increase in lifespan is calculated using the equation ~ILS=100. (T/C-1), in which T represents the average survival time in days of the experimental group and C
represents the average survival time in days of the control group.
The results obtained are illustrated in Table XVII below:
TALLE XVII

LeukaemiaAgent Route Life ~ ILS
span Control - 20 0 o.s~ NaCl s.c. 21 7 P388 VB s.c. 20 0 AFP-VB s.c. 23.8 19 control - 11.5 0 0.9~ NaCl s.c. 13 13 L1210 VB s.c. 18.8 63.5 AFP-V~3 s.c. 22 91.3 This Table XVII shows that the two types of leukaemia are treated both with ~3 alone and with the conjugate in accordance with the invention; however, the ~ILS is significantly greater for the AFP-~3 conjugate than for V~3 alone.
As regards leukaemia L1210, the effect of various concentrations of ~3 alone or of AFP 3 ~ . 21559~3 conjugate were tested. The treatment procedure is similar to that described above; however, concentrations of between 7.5 ~g and 200 ~g/kg were tested, as was the peritoneal route.
TABLE XVIII

Agent Dose Number Life % ILS
~g/kg of mice span Control0.9% NaCl 30 7 0 VB 7.5 6 8.8 25.7 AFP-VB 7.5 6 10.5 50 vn3 22 6 8.5 21.4 AFP-VB 22 6 11.3 68.6 vn3 67 6 9.2 31.4 AFP-VB 67 6 12 71.4 VB 200 6 9.7 38.6 AFP-VB 200 6 13.2 88.6 This Table XVIII shows the significantly better results obtained with the conjugates in accordance with the invention, especially at doses of 200 ~g/kg.
As emerges from the preceding text, the invention is in no way limited to that of its modes of implementation, of production and of application which have just been described in greater detail; on the contrary, it encompasses all the variants thereof which may occur to those skilled in the art, without departing from the context or the scope of the present lnventlon .

Claims (7)

- 32 -

1. Protein conjugates, of the type comprising a non-specific portion and a portion possessing a particular affinity for target cells of a specific type, characterized in that the portion possessing a particular affinity for target cells of a specific type (also referred to as substance for specific transport to the target cell) is .alpha.-foetoprotein, and in that the non-specific portion of the said conjugate is chosen from the group which comprises cytotoxic substances chosen from animal toxins, plant toxins, cytolytic enzymes and low molecular weight anti-cancer active principles and are selected from carboxyphosphamide, amboclorine, doxorubicin, bleomycin, cisplatin, vinblastine, calichemicine and methotrexate or substances which modify the metabolism of cancer cells, such as nucleotide sequences.

2. Pharmaceutical compositions, characterized in that they comprise, as active principle, at least one conjugate according to Claim 1 and at least one pharmaceutically acceptable vehicle.

3. Pharmaceutical compositions according to Claim 2, characterized in that the substances modifying the metabolism of cancer cells are chosen from nucleotide sequences which express a protein and the anti-sense nucleotide sequences.

4. Use of .alpha.-foetoprotein-cytotoxic substance protein conjugates chosen from anti-cancer active principles for the preparation of a drug intended for use in the treatment of cancers.

5. Use of .alpha.-foetoprotein-vinblastine protein conjugates in the preparation of a drug intended for use in the treatment of leukaemias.

6. Use of .alpha.-foetoprotein-doxorubicine protein conjugates in the preparation of a drug intended for use in the treatment of leukaemias.

7. Use of .alpha.-foetoprotein-calichemicine protein conjugates in the preparation of a drug intended for use in the treatment of leukaemias.